but as a structural mismatch arising from industrial transformation. Based on empirical analysis,
the study provides evidence for designing more targeted policy responses.
The analysis shows that the automotive industry has already entered a diffusion stage in terms
of SDV-related R&D investment and workforce allocation. However, the internalization of software
capabilities, which is central to SDV transformation, remains at an early stage. Although the share
of SDV-related personnel has increased, the share of core software personnel remains low. This
indicates that while SDV transition is progressing externally, the internal structural transformation
of the industry is still insufficient.
This mismatch appears more clearly across industry segments and firm sizes. Automotive and
parts manufacturing firms face high implementation burdens in SDV transition, yet they maintain a
relatively low share of software personnel, resulting in a hardware-centered transition risk. In
contrast, in the SW and service sectors, especially among mid-sized firms, bottlenecks were
observed in securing core SW personnel and generating revenue outcomes, despite relatively large
SDV workforce inputs. In terms of workforce supply and demand, the shortage rate of SW·AI
personnel was significantly higher than the overall SDV workforce shortage rate. This suggests
that the core issue is not a quantitative shortage of labor, but a qualitative mismatch across
transition stages, job functions, skill levels, and firm types.
To diagnose these issues more precisely, this study applied the Software-Centered Innovation
Capability Index (SDICI). SDICI comprehensively measures a firm’s strategic, technological, human,
business, and ecosystem capabilities. Using this index, the study analyzed the relationship between
SW·AI workforce structures and firm performance. The results show that SDV transition is not a
linear process in which capabilities improve gradually, but a non-linear transformation process in
which bottlenecks change around certain threshold points. In the early stage, basic capabilities and
implementation foundations act as major constraints. As maturity increases, the bottlenecks shift
toward investment expansion, advanced workforce acquisition, standardization, external
collaboration, business model expansion, and performance transfer.
Cluster analysis classified firms into four types according to their SDV transition level:
foundation-building, implementation and transition, expansion and growth, and leading and diffusion.
Each group has different bottlenecks and policy needs, indicating that future policies should be
designed in a differentiated manner according to firms’ maturity levels. Regression analysis
further confirmed that technological capability is a key factor explaining both the generation and
expansion of SDV-related revenue. Human capability, meanwhile, was found to amplify
performance particularly in the scale-up stage, after a certain level of technological foundation